Currently, blkg's are embedded in private data blkcg policy private
data structure and thus allocated and freed by policies. This leads
to duplicate codes in policies, hinders implementing common part in
blkcg core with strong semantics, and forces duplicate blkg's for the
same cgroup-q association.
This patch introduces struct blkg_policy_data which is a separate data
structure chained from blkg. Policies specifies the amount of private
data it needs in its blkio_policy_type->pdata_size and blkcg core
takes care of allocating them along with blkg which can be accessed
using blkg_to_pdata(). blkg can be determined from pdata using
pdata_to_blkg(). blkio_alloc_group_fn() method is accordingly updated
to blkio_init_group_fn().
For consistency, tg_of_blkg() and cfqg_of_blkg() are replaced with
blkg_to_tg() and blkg_to_cfqg() respectively, and functions to map in
the reverse direction are added.
Except that policy specific data now lives in a separate data
structure from blkg, this patch doesn't introduce any functional
difference.
This will be used to unify blkg's for different policies.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently, blkg points to the associated blkcg via its css_id. This
unnecessarily complicates dereferencing blkcg. Let blkg hold a
reference to the associated blkcg and point directly to it and disable
css_id on blkio_subsys.
This change requires splitting blkiocg_destroy() into
blkiocg_pre_destroy() and blkiocg_destroy() so that all blkg's can be
destroyed and all the blkcg references held by them dropped during
cgroup removal.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blkg->dev is dev_t recording the device number of the block device for
the associated request_queue. It is used to identify the associated
block device when printing out configuration or stats.
This is redundant to begin with. A blkg is an association between a
cgroup and a request_queue and it of course is possible to reach
request_queue from blkg and synchronization conventions are in place
for safe q dereferencing, so this shouldn't be necessary from the
beginning. Furthermore, it's initialized by sscanf()ing the device
name of backing_dev_info. The mind boggles.
Anyways, if blkg is visible under rcu lock, we *know* that the
associated request_queue hasn't gone away yet and its bdi is
registered and alive - blkg can't be created for request_queue which
hasn't been fully initialized and it can't go away before blkg is
removed.
Let stat and conf read functions get device name from
blkg->q->backing_dev_info.dev and pass it down to printing functions
and remove blkg->dev.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blkcg is very peculiar in that it allows setting and remembering
configurations for non-existent devices by maintaining separate data
structures for configuration.
This behavior is completely out of the usual norms and outright
confusing; furthermore, it uses dev_t number to match the
configuration to devices, which is unpredictable to begin with and
becomes completely unuseable if EXT_DEVT is fully used.
It is wholely unnecessary - we already have fully functional userland
mechanism to program devices being hotplugged which has full access to
device identification, connection topology and filesystem information.
Add a new struct blkio_group_conf which contains all blkcg
configurations to blkio_group and let blkio_group, which can be
created iff the associated device exists and is removed when the
associated device goes away, carry all configurations.
Note that, after this patch, all newly created blkg's will always have
the default configuration (unlimited for throttling and blkcg's weight
for propio).
This patch makes blkio_policy_node meaningless but doesn't remove it.
The next patch will.
-v2: Updated to retry after short sleep if blkg lookup/creation failed
due to the queue being temporarily bypassed as indicated by
-EBUSY return. Pointed out by Vivek.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Cc: Kay Sievers <kay.sievers@vrfy.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently both blk-throttle and cfq-iosched implement their own
blkio_group creation code in throtl_get_tg() and cfq_get_cfqg(). This
patch factors out the common code into blkg_lookup_create(), which
returns ERR_PTR value so that transitional failures due to queue
bypass can be distinguished from other failures.
* New plkio_policy_ops methods blkio_alloc_group_fn() and
blkio_link_group_fn added. Both are transitional and will be
removed once the blkg management code is fully moved into
blk-cgroup.c.
* blkio_alloc_group_fn() allocates policy-specific blkg which is
usually a larger data structure with blkg as the first entry and
intiailizes it. Note that initialization of blkg proper, including
percpu stats, is responsibility of blk-cgroup proper.
Note that default config (weight, bps...) initialization is done
from this method; otherwise, we end up violating locking order
between blkcg and q locks via blkcg_get_CONF() functions.
* blkio_link_group_fn() is called under queue_lock and responsible for
linking the blkg to the queue. blkcg side is handled by blk-cgroup
proper.
* The common blkg creation function is named blkg_lookup_create() and
blkiocg_lookup_group() is renamed to blkg_lookup() for consistency.
Also, throtl / cfq related functions are similarly [re]named for
consistency.
This simplifies blkcg policy implementations and enables further
cleanup.
-v2: Vivek noticed that blkg_lookup_create() incorrectly tested
blk_queue_dead() instead of blk_queue_bypass() leading a user of
the function ending up creating a new blkg on bypassing queue.
This is a bug introduced while relocating bypass patches before
this one. Fixed.
-v3: ERR_PTR patch folded into this one. @for_root added to
blkg_lookup_create() to allow creating root group on a bypassed
queue during elevator switch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
For root blkg, blk_throtl_init() was using throtl_alloc_tg()
explicitly and cfq_init_queue() was manually initializing embedded
cfqd->root_group, adding unnecessarily different code paths to blkg
handling.
Make both use the usual blkio_group get functions - throtl_get_tg()
and cfq_get_cfqg() - for the root blkio_group too. Note that
blk_throtl_init() callsite is pushed downwards in
blk_alloc_queue_node() so that @q is sufficiently initialized for
throtl_get_tg().
This simplifies root blkg handling noticeably for cfq and will allow
further modularization of blkcg API.
-v2: Vivek pointed out that using cfq_get_cfqg() won't work if
CONFIG_CFQ_GROUP_IOSCHED is disabled. Fix it by factoring out
initialization of base part of cfqg into cfq_init_cfqg_base() and
alloc/init/free explicitly if !CONFIG_CFQ_GROUP_IOSCHED.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blkgio_group is association between a block cgroup and a queue for a
given policy. Using opaque void * for association makes things
confusing and hinders factoring of common code. Use request_queue *
and, if necessary, policy id instead.
This will help block cgroup API cleanup.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In both blkg get functions - throtl_get_tg() and cfq_get_cfqg(),
instead of obtaining blkcg of %current explicitly, let the caller
specify the blkcg to use as parameter and make both functions hold on
to the blkcg.
This is part of block cgroup interface cleanup and will help making
blkcg API more modular.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
rcu_read_lock() in throtl_get_tb() and cfq_get_cfqg() holds onto
@blkcg while looking up blkg. For API cleanup, the next patch will
make the caller responsible for determining @blkcg to look blkg from
and let them specify it as a parameter. Move rcu read locking out to
the callers to prepare for the change.
-v2: Originally this patch was described as a fix for RCU read locking
bug around @blkg, which Vivek pointed out to be incorrect. It
was from misunderstanding the role of rcu locking as protecting
@blkg not @blkcg. Patch description updated.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Elevator switch may involve changes to blkcg policies. Implement
shoot down of blkio_groups.
Combined with the previous bypass updates, the end goal is updating
blkcg core such that it can ensure that blkcg's being affected become
quiescent and don't have any per-blkg data hanging around before
commencing any policy updates. Until queues are made aware of the
policies that applies to them, as an interim step, all per-policy blkg
data will be shot down.
* blk-throtl doesn't need this change as it can't be disabled for a
live queue; however, update it anyway as the scheduled blkg
unification requires this behavior change. This means that
blk-throtl configuration will be unnecessarily lost over elevator
switch. This oddity will be removed after blkcg learns to associate
individual policies with request_queues.
* blk-throtl dosen't shoot down root_tg. This is to ease transition.
Unified blkg will always have persistent root group and not shooting
down root_tg for now eases transition to that point by avoiding
having to update td->root_tg and is safe as blk-throtl can never be
disabled
-v2: Vivek pointed out that group list is not guaranteed to be empty
on return from clear function if it raced cgroup removal and
lost. Fix it by waiting a bit and retrying. This kludge will
soon be removed once locking is updated such that blkg is never
in limbo state between blkcg and request_queue locks.
blk-throtl no longer shoots down root_tg to avoid breaking
td->root_tg.
Also, Nest queue_lock inside blkio_list_lock not the other way
around to avoid introduce possible deadlock via blkcg lock.
-v3: blkcg_clear_queue() repositioned and renamed to
blkg_destroy_all() to increase consistency with later changes.
cfq_clear_queue() updated to check q->elevator before
dereferencing it to avoid NULL dereference on not fully
initialized queues (used by later change).
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
elevator_ops->elevator_init_fn() has a weird return value. It returns
a void * which the caller should assign to q->elevator->elevator_data
and %NULL return denotes init failure.
Update such that it returns integer 0/-errno and sets elevator_data
directly as necessary.
This makes the interface more conventional and eases further cleanup.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
cfq has been registering zeroed blkio_poilcy_cfq if CFQ_GROUP_IOSCHED
is disabled. This fortunately doesn't collide with blk-throtl as
BLKIO_POLICY_PROP is zero but is unnecessary and risky. Just don't
register it if not enabled.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
icq->changed was used for ICQ_*_CHANGED bits. Rename it to flags and
access it under ioc->lock instead of using atomic bitops.
ioc_get_changed() is added so that the changed part can be fetched and
cleared as before.
icq->flags will be used to carry other flags.
Signed-off-by: Tejun Heo <tj@kernel.org>
Tested-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Plug merge calls two elevator callbacks outside queue lock -
elevator_allow_merge_fn() and elevator_bio_merged_fn(). Although
attempt_plug_merge() suggests that elevator is guaranteed to be there
through the existing request on the plug list, nothing prevents plug
merge from calling into dying or initializing elevator.
For regular merges, bypass ensures elvpriv count to reach zero, which
in turn prevents merges as all !ELVPRIV requests get REQ_SOFTBARRIER
from forced back insertion. Plug merge doesn't check ELVPRIV, and, as
the requests haven't gone through elevator insertion yet, it doesn't
have SOFTBARRIER set allowing merges on a bypassed queue.
This, for example, leads to the following crash during elevator
switch.
BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
IP: [<ffffffff813b34e9>] cfq_allow_merge+0x49/0xa0
PGD 112cbc067 PUD 115d5c067 PMD 0
Oops: 0000 [#1] PREEMPT SMP
CPU 1
Modules linked in: deadline_iosched
Pid: 819, comm: dd Not tainted 3.3.0-rc2-work+ #76 Bochs Bochs
RIP: 0010:[<ffffffff813b34e9>] [<ffffffff813b34e9>] cfq_allow_merge+0x49/0xa0
RSP: 0018:ffff8801143a38f8 EFLAGS: 00010297
RAX: 0000000000000000 RBX: ffff88011817ce28 RCX: ffff880116eb6cc0
RDX: 0000000000000000 RSI: ffff880118056e20 RDI: ffff8801199512f8
RBP: ffff8801143a3908 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000001 R11: 0000000000000000 R12: ffff880118195708
R13: ffff880118052aa0 R14: ffff8801143a3d50 R15: ffff880118195708
FS: 00007f19f82cb700(0000) GS:ffff88011fc80000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
CR2: 0000000000000008 CR3: 0000000112c6a000 CR4: 00000000000006e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process dd (pid: 819, threadinfo ffff8801143a2000, task ffff880116eb6cc0)
Stack:
ffff88011817ce28 ffff880118195708 ffff8801143a3928 ffffffff81391bba
ffff88011817ce28 ffff880118195708 ffff8801143a3948 ffffffff81391bf1
ffff88011817ce28 0000000000000000 ffff8801143a39a8 ffffffff81398e3e
Call Trace:
[<ffffffff81391bba>] elv_rq_merge_ok+0x4a/0x60
[<ffffffff81391bf1>] elv_try_merge+0x21/0x40
[<ffffffff81398e3e>] blk_queue_bio+0x8e/0x390
[<ffffffff81396a5a>] generic_make_request+0xca/0x100
[<ffffffff81396b04>] submit_bio+0x74/0x100
[<ffffffff811d45c2>] __blockdev_direct_IO+0x1ce2/0x3450
[<ffffffff811d0dc7>] blkdev_direct_IO+0x57/0x60
[<ffffffff811460b5>] generic_file_aio_read+0x6d5/0x760
[<ffffffff811986b2>] do_sync_read+0xe2/0x120
[<ffffffff81199345>] vfs_read+0xc5/0x180
[<ffffffff81199501>] sys_read+0x51/0x90
[<ffffffff81aeac12>] system_call_fastpath+0x16/0x1b
There are multiple ways to fix this including making plug merge check
ELVPRIV; however,
* Calling into elevator outside queue lock is confusing and
error-prone.
* Requests on plug list aren't known to the elevator. They aren't on
the elevator yet, so there's no elevator specific state to update.
* Given the nature of plug merges - collecting bio's for the same
purpose from the same issuer - elevator specific restrictions aren't
applicable.
So, simply don't call into elevator methods from plug merge by moving
elv_bio_merged() from bio_attempt_*_merge() to blk_queue_bio(), and
using blk_try_merge() in attempt_plug_merge().
This is based on Jens' patch to skip elevator_allow_merge_fn() from
plug merge.
Note that this makes per-cgroup merged stats skip plug merging.
Signed-off-by: Tejun Heo <tj@kernel.org>
LKML-Reference: <4F16F3CA.90904@kernel.dk>
Original-patch-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
put_io_context() performed a complex trylock dancing to avoid
deferring ioc release to workqueue. It was also broken on UP because
trylock was always assumed to succeed which resulted in unbalanced
preemption count.
While there are ways to fix the UP breakage, even the most
pathological microbench (forced ioc allocation and tight fork/exit
loop) fails to show any appreciable performance benefit of the
optimization. Strip it out. If there turns out to be workloads which
are affected by this change, simpler optimization from the discussion
thread can be applied later.
Signed-off-by: Tejun Heo <tj@kernel.org>
LKML-Reference: <1328514611.21268.66.camel@sli10-conroe>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
cfq_slice_expired will change saved_workload_slice. It should be called
first so saved_workload_slice is correctly set to 0 after workload type
is changed.
This fixes the code order changed by 54b466e44b.
Tested-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
All requests of a queue could be merged to other requests of other queue.
Such queue will not have request in it, but it's in service tree. This
will cause kernel oops.
I encounter a BUG_ON() in cfq_dispatch_request() with next patch, but the
issue should exist without the patch.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now block layer knows everything necessary to create and associate
icq's with requests. Move ioc_create_icq() to blk-ioc.c and update
get_request() such that, if elevator_type->icq_size is set, requests
are automatically associated with their matching icq's before
elv_set_request(). io_context reference is also managed by block core
on request alloc/free.
* Only ioprio/cgroup changed handling remains from cfq_get_cic().
Collapsed into cfq_set_request().
* This removes queue kicking on icq allocation failure (for now). As
icq allocation failure is rare and the only effect of queue kicking
achieved was possibily accelerating queue processing, this change
shouldn't be noticeable.
There is a larger underlying problem. Unlike request allocation,
icq allocation is not guaranteed to succeed eventually after
retries. The number of icq is unbound and thus mempool can't be the
solution either. This effectively adds allocation dependency on
memory free path and thus possibility of deadlock.
This usually wouldn't happen because icq allocation is not a hot
path and, even when the condition triggers, it's highly unlikely
that none of the writeback workers already has icq.
However, this is still possible especially if elevator is being
switched under high memory pressure, so we better get it fixed.
Probably the only solution is just bypassing elevator and appending
to dispatch queue on any elevator allocation failure.
* Comment added to explain how icq's are managed and synchronized.
This completes cleanup of io_context interface.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add elevator_ops->elevator_init_icq_fn() and restructure
cfq_create_cic() and rename it to ioc_create_icq().
The new function expects its caller to pass in io_context, uses
elevator_type->icq_cache, handles generic init, calls the new elevator
operation for elevator specific initialization, and returns pointer to
created or looked up icq. This leaves cfq_icq_pool variable without
any user. Removed.
This prepares for io_context interface cleanup and doesn't introduce
any functional difference.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
With kmem_cache managed by blk-ioc, io_cq exit/release can be moved to
blk-ioc too. The odd ->io_cq->exit/release() callbacks are replaced
with elevator_ops->elevator_exit_icq_fn() with unlinking from both ioc
and q, and freeing automatically handled by blk-ioc. The elevator
operation only need to perform exit operation specific to the elevator
- in cfq's case, exiting the cfqq's.
Also, clearing of io_cq's on q detach is moved to block core and
automatically performed on elevator switch and q release.
Because the q io_cq points to might be freed before RCU callback for
the io_cq runs, blk-ioc code should remember to which cache the io_cq
needs to be freed when the io_cq is released. New field
io_cq->__rcu_icq_cache is added for this purpose. As both the new
field and rcu_head are used only after io_cq is released and the
q/ioc_node fields aren't, they are put into unions.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Let elevators set ->icq_size and ->icq_align in elevator_type and
elv_register() and elv_unregister() respectively create and destroy
kmem_cache for icq.
* elv_register() now can return failure. All callers updated.
* icq caches are automatically named "ELVNAME_io_cq".
* cfq_slab_setup/kill() are collapsed into cfq_init/exit().
* While at it, minor indentation change for iosched_cfq.elevator_name
for consistency.
This will help moving icq management to block core. This doesn't
introduce any functional change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that all io_cq related data structures are in block core layer,
io_cq lookup can be moved from cfq-iosched.c to blk-ioc.c.
Lookup logic from cfq_cic_lookup() is moved to ioc_lookup_icq() with
parameter return type changes (cfqd -> request_queue, cfq_io_cq ->
io_cq) and cfq_cic_lookup() becomes thin wrapper around
cfq_cic_lookup().
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Most of icq management is about to be moved out of cfq into blk-ioc.
This patch prepares for it.
* Move cfqd->icq_list to request_queue->icq_list
* Make request explicitly point to icq instead of through elevator
private data. ->elevator_private[3] is replaced with sub struct elv
which contains icq pointer and priv[2]. cfq is updated accordingly.
* Meaningless clearing of ->elevator_private[0] removed from
elv_set_request(). At that point in code, the field was guaranteed
to be %NULL anyway.
This patch doesn't introduce any functional change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently io_context and cfq logics are mixed without clear boundary.
Most of io_context is independent from cfq but cfq_io_context handling
logic is dispersed between generic ioc code and cfq.
cfq_io_context represents association between an io_context and a
request_queue, which is a concept useful outside of cfq, but it also
contains fields which are useful only to cfq.
This patch takes out generic part and put it into io_cq (io
context-queue) and the rest into cfq_io_cq (cic moniker remains the
same) which contains io_cq. The following changes are made together.
* cfq_ttime and cfq_io_cq now live in cfq-iosched.c.
* All related fields, functions and constants are renamed accordingly.
* ioc->ioc_data is now "struct io_cq *" instead of "void *" and
renamed to icq_hint.
This prepares for io_context API cleanup. Documentation is currently
sparse. It will be added later.
Changes in this patch are mechanical and don't cause functional
change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When called under queue_lock, current_io_context() triggers lockdep
warning if it hits allocation path. This is because io_context
installation is protected by task_lock which is not IRQ safe, so it
triggers irq-unsafe-lock -> irq -> irq-safe-lock -> irq-unsafe-lock
deadlock warning.
Given the restriction, accessor + creator rolled into one doesn't work
too well. Drop current_io_context() and let the users access
task->io_context directly inside queue_lock combined with explicit
creation using create_io_context().
Future ioc updates will further consolidate ioc access and the create
interface will be unexported.
While at it, relocate ioc internal interface declarations in blk.h and
add section comments before and after.
This patch does not introduce functional change.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that lazy paths are removed, cfqd_dead_key() is meaningless and
cic->q can be used whereever cic->key is used. Kill cic->key.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that cic's are immediately unlinked under both locks, there's no
need to count and drain cic's before module unload. RCU callback
completion is waited with rcu_barrier().
While at it, remove residual RCU operations on cic_list.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that all cic's are immediately unlinked from both ioc and queue,
lazy dropping from lookup path and trimming on elevator unregister are
unnecessary. Kill them and remove now unused elevator_ops->trim().
This also leaves call_for_each_cic() without any user. Removed.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
cic is association between io_context and request_queue. A cic is
linked from both ioc and q and should be destroyed when either one
goes away. As ioc and q both have their own locks, locking becomes a
bit complex - both orders work for removal from one but not from the
other.
Currently, cfq tries to circumvent this locking order issue with RCU.
ioc->lock nests inside queue_lock but the radix tree and cic's are
also protected by RCU allowing either side to walk their lists without
grabbing lock.
This rather unconventional use of RCU quickly devolves into extremely
fragile convolution. e.g. The following is from cfqd going away too
soon after ioc and q exits raced.
general protection fault: 0000 [#1] PREEMPT SMP
CPU 2
Modules linked in:
[ 88.503444]
Pid: 599, comm: hexdump Not tainted 3.1.0-rc10-work+ #158 Bochs Bochs
RIP: 0010:[<ffffffff81397628>] [<ffffffff81397628>] cfq_exit_single_io_context+0x58/0xf0
...
Call Trace:
[<ffffffff81395a4a>] call_for_each_cic+0x5a/0x90
[<ffffffff81395ab5>] cfq_exit_io_context+0x15/0x20
[<ffffffff81389130>] exit_io_context+0x100/0x140
[<ffffffff81098a29>] do_exit+0x579/0x850
[<ffffffff81098d5b>] do_group_exit+0x5b/0xd0
[<ffffffff81098de7>] sys_exit_group+0x17/0x20
[<ffffffff81b02f2b>] system_call_fastpath+0x16/0x1b
The only real hot path here is cic lookup during request
initialization and avoiding extra locking requires very confined use
of RCU. This patch makes cic removal from both ioc and request_queue
perform double-locking and unlink immediately.
* From q side, the change is almost trivial as ioc->lock nests inside
queue_lock. It just needs to grab each ioc->lock as it walks
cic_list and unlink it.
* From ioc side, it's a bit more difficult because of inversed lock
order. ioc needs its lock to walk its cic_list but can't grab the
matching queue_lock and needs to perform unlock-relock dancing.
Unlinking is now wholly done from put_io_context() and fast path is
optimized by using the queue_lock the caller already holds, which is
by far the most common case. If the ioc accessed multiple devices,
it tries with trylock. In unlikely cases of fast path failure, it
falls back to full double-locking dance from workqueue.
Double-locking isn't the prettiest thing in the world but it's *far*
simpler and more understandable than RCU trick without adding any
meaningful overhead.
This still leaves a lot of now unnecessary RCU logics. Future patches
will trim them.
-v2: Vivek pointed out that cic->q was being dereferenced after
cic->release() was called. Updated to use local variable @this_q
instead.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
* cfq_cic_lookup() may be called without queue_lock and multiple tasks
can execute it simultaneously for the same shared ioc. Nothing
prevents them racing each other and trying to drop the same dead cic
entry multiple times.
* smp_wmb() in cfq_exit_cic() doesn't really do anything and nothing
prevents cfq_cic_lookup() seeing stale cic->key. This usually
doesn't blow up because by the time cic is exited, all requests have
been drained and new requests are terminated before going through
elevator. However, it can still be triggered by plug merge path
which doesn't grab queue_lock and thus can't check DEAD state
reliably.
This patch updates lookup locking such that,
* Lookup is always performed under queue_lock. This doesn't add any
more locking. The only issue is cfq_allow_merge() which can be
called from plug merge path without holding any lock. For now, this
is worked around by using cic of the request to merge into, which is
guaranteed to have the same ioc. For longer term, I think it would
be best to separate out plug merge method from regular one.
* Spurious ioc->lock locking around cic lookup hint assignment
dropped.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
cfq_get_io_context() would fail if multiple tasks race to insert cic's
for the same association. This patch restructures
cfq_get_io_context() such that slow path insertion race is handled
properly.
Note that the restructuring also makes cfq_get_io_context() called
under queue_lock and performs both ioc and cfqd insertions while
holding both ioc and queue locks. This is part of on-going locking
tightening and will be used to simplify synchronization rules.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
ioprio/cgroup change was handled by marking the changed state in ioc
and, on the following access to the ioc, performing RCU-protected
iteration through all cic's grabbing the matching queue_lock.
This patch moves the changed state to each cic. When ioprio or cgroup
changes, the respective bit is set on all cic's of the ioc and when
each of those cic (not ioc) is accessed, change is applied for that
specific ioc-queue pair.
This also fixes the following two race conditions between setting and
clearing of changed states.
* Missing barrier between assign/load of ioprio and ioprio_changed
allowed applying old ioprio.
* Change requests could happen between application of change and
clearing of changed variables.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Make the following changes to prepare for ioc/cic management cleanup.
* Add cic->q so that ioc can determine the associated queue without
querying cfq. This will eventually replace ->key.
* Factor out cfq_release_cic() from cic_free_func(). This function
assumes that the caller handled locking.
* Rename __cfq_exit_single_io_context() to cfq_exit_cic() and make it
take only @cic.
* Restructure cfq_cic_link() for future updates.
This patch doesn't introduce any functional changes.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Ignoring copy_io() during fork, io_context can be allocated from two
places - current_io_context() and set_task_ioprio(). The former is
always called from local task while the latter can be called from
different task. The synchornization between them are peculiar and
dubious.
* current_io_context() doesn't grab task_lock() and assumes that if it
saw %NULL ->io_context, it would stay that way until allocation and
assignment is complete. It has smp_wmb() between alloc/init and
assignment.
* set_task_ioprio() grabs task_lock() for assignment and does
smp_read_barrier_depends() between "ioc = task->io_context" and "if
(ioc)". Unfortunately, this doesn't achieve anything - the latter
is not a dependent load of the former. ie, if ioc itself were being
dereferenced "ioc->xxx", it would mean something (not sure what tho)
but as the code currently stands, the dependent read barrier is
noop.
As only one of the the two test-assignment sequences is task_lock()
protected, the task_lock() can't do much about race between the two.
Nothing prevents current_io_context() and set_task_ioprio() allocating
its own ioc for the same task and overwriting the other's.
Also, set_task_ioprio() can race with exiting task and create a new
ioc after exit_io_context() is finished.
ioc get/put doesn't have any reason to be complex. The only hot path
is accessing the existing ioc of %current, which is simple to achieve
given that ->io_context is never destroyed as long as the task is
alive. All other paths can happily go through task_lock() like all
other task sub structures without impacting anything.
This patch updates ioc get/put so that it becomes more conventional.
* alloc_io_context() is replaced with get_task_io_context(). This is
the only interface which can acquire access to ioc of another task.
On return, the caller has an explicit reference to the object which
should be put using put_io_context() afterwards.
* The functionality of current_io_context() remains the same but when
creating a new ioc, it shares the code path with
get_task_io_context() and always goes through task_lock().
* get_io_context() now means incrementing ref on an ioc which the
caller already has access to (be that an explicit refcnt or implicit
%current one).
* PF_EXITING inhibits creation of new io_context and once
exit_io_context() is finished, it's guaranteed that both ioc
acquisition functions return %NULL.
* All users are updated. Most are trivial but
smp_read_barrier_depends() removal from cfq_get_io_context() needs a
bit of explanation. I suppose the original intention was to ensure
ioc->ioprio is visible when set_task_ioprio() allocates new
io_context and installs it; however, this wouldn't have worked
because set_task_ioprio() doesn't have wmb between init and install.
There are other problems with this which will be fixed in another
patch.
* While at it, use NUMA_NO_NODE instead of -1 for wildcard node
specification.
-v2: Vivek spotted contamination from debug patch. Removed.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
cfq allocates per-queue id using ida and uses it to index cic radix
tree from io_context. Move it to q->id and allocate on queue init and
free on queue release. This simplifies cfq a bit and will allow for
further improvements of io context life-cycle management.
This patch doesn't introduce any functional difference.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a new REQ_PRIO to let requests preempt others in the cfq I/O schedule,
and lave REQ_META purely for marking requests as metadata in blktrace.
All existing callers of REQ_META except for XFS are updated to also
set REQ_PRIO for now.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Namhyung Kim <namhyung@gmail.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
We have a kernel build regression since 3.1-rc1, which is about 10%
regression. The kernel source is in an ext3 filesystem.
Alex Shi bisect it to commit:
commit a07405b780
Author: Justin TerAvest <teravest@google.com>
Date: Sun Jul 10 22:09:19 2011 +0200
cfq: Remove special treatment for metadata rqs.
Apparently this is caused by lack metadata preemption, where ext3/ext4
do use READ_META. I didn't see a way to fix the issue, so suggest
reverting the patch.
This reverts commit a07405b780.
Reported-by: Alex Shi<alex.shi@intel.com>
Reported-by: Shaohua Li<shaohua.li@intel.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
FQ keeps track of number of groups which are linked on blkcg->blkg_list.
This is useful to avoid races between queue exit and cgroup exit code
paths. So if at the request queue exit time linked group count is not
zero, that means there are some group out there which is yet to be
deleted under rcu read period and queue exit code should wait for
on rcu period.
In my previous patch I forgot to decrease the number of group count.
So in current form, we nr_blkcg_linked_grps is always non-zero and
we will always wait one rcu period (if BLK_CGROUP=y). The side effect
of this is that it can increase boot time. I am surprised, nobody
complained so far.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Currently when the last queue of a group has no request, we don't expire
the queue to hope request from the group comes soon, so the group doesn't
miss its share. But if the think time is big, the assumption isn't correct
and we just waste bandwidth. In such case, we don't do idle.
[global]
runtime=30
direct=1
[test1]
cgroup=test1
cgroup_weight=1000
rw=randread
ioengine=libaio
size=500m
runtime=30
directory=/mnt
filename=file1
thinktime=9000
[test2]
cgroup=test2
cgroup_weight=1000
rw=randread
ioengine=libaio
size=500m
runtime=30
directory=/mnt
filename=file2
patched base
test1 64k 39k
test2 548k 540k
total 604k 578k
group1 gets much better throughput because it waits less time.
To check if the patch changes behavior of queue without think time. I also
tried to give test1 2ms think time or no think time. The test result is stable.
The thoughput doesn't change with/without the patch.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Currently when the last queue of a service tree has no request, we don't
expire the queue to hope request from the service tree comes soon, so the
service tree doesn't miss its share. But if the think time is big, the
assumption isn't correct and we just waste bandwidth. In such case, we
don't do idle.
[global]
runtime=10
direct=1
[test1]
rw=randread
ioengine=libaio
size=500m
directory=/mnt
filename=file1
thinktime=9000
[test2]
rw=read
ioengine=libaio
size=1G
directory=/mnt
filename=file2
patched base
test1 41k/s 33k/s
test2 15868k/s 15789k/s
total 15902k/s 15817k/s
A slightly better
To check if the patch changes behavior of queue without think time. I also
tried to give test1 2ms think time or no think time. The test has variation
even without the patch, but the average throughput doesn't change with/without
the patch.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Move the variables to do think time check to a sepatate struct. This is
to prepare adding think time check for service tree and group. No
functional change.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Acked-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
fs_excl is a poor man's priority inheritance for filesystems to hint to
the block layer that an operation is important. It was never clearly
specified, not widely adopted, and will not prevent starvation in many
cases (like across cgroups).
fs_excl was introduced with the time sliced CFQ IO scheduler, to
indicate when a process held FS exclusive resources and thus needed
a boost.
It doesn't cover all file systems, and it was never fully complete.
Lets kill it.
Signed-off-by: Justin TerAvest <teravest@google.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
There is no consistency among filesystems from what bios (or requests)
are marked as being metadata. It's interesting to expose this in traces,
but we shouldn't schedule the requests differently based on whether or
not they're marked as being metadata.
Signed-off-by: Justin TerAvest <teravest@google.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
ioc->ioc_data is rcu protectd, so uses correct API to access it.
This doesn't change any behavior, but just make code consistent.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: stable@kernel.org # after ab4bd22d
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
I got a rcu warnning at boot. the ioc->ioc_data is rcu_deferenced, but
doesn't hold rcu_read_lock.
Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Cc: stable@kernel.org # after ab4bd22d
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Use the compiler to verify format strings and arguments.
Fix fallout.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Use the compiler to verify format strings and arguments.
Fix fallout.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Tejun Heo